Designing the “right” thing for the “right” persona

How to help your clients prioritize personas for better design decisions

I applied my first, ‘official’ personas back when I worked at the University of Houston. When I designed the Distance Education website, I focused on the New Student persona, and the student services office call volume fell by 10% despite a 20% increase in enrollment. When you prioritize the right user’s needs, you can see a massive affect on the success of your experience design.

When you prioritize the right user’s needs, you can see a massive affect on the success of your experience design.

Back then, I lucked out. No one understood what I did. I built the entire site, from concept to code. I, alone, decided we should design the site for the New Student. Nowadays, my projects keep getting bigger, more and more people get involved. Now, when choosing a primary persona, I facilitate that decision in a room full of stakeholders who don’t see eye to eye.

Why prioritized personas?

Despite all our fancy technology and methodology, most experiences are pretty simple. It’s not that we can’t create insane, high-fidelity, crazy-complex, and super-personalized experiences. We can. But they’re hard to plan, hard to produce, and harder to maintain. It’s not that we don’t have the right technology or the right methodology. We don’t have the mental space in our little heads to consume these kinds of problems.

In my practice, I have stakeholders prioritize everything. Everything. When they prioritize, they agree that Thing One is more important than Thing Two. It’s that simple.

Sometimes we have an explicit reason for the prioritization. Other times, it’s just a gut feeling. That’s ok. Once you have a priority, it makes decisions easier. You spend less time deciding simple things, so you have more head space for the complex things. You get to spend your head space on the parts of the experience that make it awesome.

Back at the University of Houston, when I prioritized the needs of the New Student over the needs of the Returning Student or the Faculty Member, it became easier to decide what went at the top of the home page. It made it easier to decide what labels to use to describe classes. It’s harder to create a label that is equally understood by Faculty, Returning Students, AND New Students. It’s easier to choose a label that’s well understood by one group, and learnable by other groups. Prioritization makes all kinds of design decisions a lot easier.

You can prioritize personas any one of a million ways. I have four ways I like to use:

1. Population, what percentage of users does a persona represent?
2. Frequency, what percentage of total interactions does a single user group engage in?
3. Criticality, how important is a single user group’s interaction?
4. Influence, how much influence does one user have on others?

One or more of these criteria should apply to any project you work on.

Prioritizing personas with single criteria

It’s pretty straightforward to prioritize personas using criteria like population:

1. Take a look at your personas
2. For each persona, calculate or estimate the percentage of the total user population that persona represents
3. Rank your personas from the highest to the lowest population

The most important persona could be the one that represents the largest group of users.

Prioritizing by population makes a lot of sense. It’s reasonable to assume your largest user groups need more design love.

To rank your personas, you can use any quantity. For population, I often use percentages. For more abstract criteria like criticality or influence I use a scale from one to three or high, medium, low. As long as you can use a single value to rank the persona, you can prioritize the persona.

You can use any of those four criteria in the same way. Do 20% of your users create 80% of your sales, support calls, or recommendations? Then you might prioritize by impact.

Influence is a common factor on B2B sites. Your primary buyer might be someone in middle management, but your site spends an inordinate amount of time speaking to executives, because when they go look at your product, their perception carries a huge amount of influence on whether or not they will spend the money. Strategic importance is similar to influence. Is there one user group that you’ve identified as your primary target market?

For isolated, one-dimensional experiences, prioritizing with one criteria might be enough, but you’re probably not that lucky.

Prioritizing personas with multiple criteria

On many sites, it’s pretty easy to apply multiple criteria. For example, it’s pretty common for me to rank personas across three criteria like, Population, Frequency, and Criticality. To prioritize with more than one criteria, you want to make it easy for people to understand your multi-dimensional ranking:

1. Identify the most relevant criteria
2. List those criteria from most important to least important
3. For each persona, calculate their score for each criteria

For example, on the Distance Education site, I might have scored the New Student like this:

New Student

• Population: 20%
• Frequency: Weekly
• Criticality: High

If you notice, the above scoring doesn’t actually prioritize the New Student persona above the Returning Student persona. In order to prioritize, you have to compare multiple personas. You have to evaluate the Returning Student with the same criteria yo used to evaluate the New Student.

Returning Student

• Population: 80%
• Frequency: Monthly
• Criticality: Low

You can put these attributes into your favorite/least dreaded spreadsheet and it makes it easier to compare each persona.

Documenting your personas and their attributes in Excel makes it easy to compare them to one another.

To sort by multiple criteria, we’ll change the low, medium, high ranking we used for criticality to numbers. I substitute them with 3, 2, and 1 for high, medium, and low.

Converting attributes to numbers makes it easy to sort them. When you sort by population first, then the Returning User persona is obviously more important.

Once you’re in a spreadsheet, you can sort on multiple criteria. You can sort your personas from high to low by population, then by frequency, then by criticality. (This is why step two above is to list your criteria from most to least important.)

Most spreadsheet programs allow you to sort by multiple columns, in a priority order.

But what if you can’t decide what criteria is most important? Worse, what if the politics on your project are so bad that you don’t want to decide? What if you need one clear number that communicates what persona is more important.

Prioritizing personas with a blended metric like ‘Engagement Volume’

When you calculate the volume of a three-dimensional object, you multiply the length by the width by the height. That’s the object’s volume.

Similarly, if you have given a persona’s rankings across several criteria, you can multiply the scores together to create a single, blended score. It’s like that persona’s “volume”. And once you have a single, blended score, you can compare one persona to another along one dimension instead of multiple dimensions.

New Student

• Population: 20%
• Frequency: 3, for high
• Criticality: 3, for high
• Total Engagement Volume: 180 (20 x 3 x 3 = 180)

Returning Student

• Population: 80%
• Frequency: 2, for medium
• Criticality: 1, for low
• Total Engagement Volume: 160 (80 x 2 x 1 = 160)

The total engagement volume for the New Student is 180 while the total engagement volume for the Returning Student is 160.

Which persona is more important? The New Student. Not because you’re biased, but because you’ve analyzed each persona using three criteria and identified which persona fits best. Unlike in the sorting example we saw above, the New Student is more important than the Returning Student because although population is our most important criteria, when you take into account all criteria, the New Student is more important.

160 and 180 are odd numbers. It can be useful to evaluate everything using 1, 2, and 3 for low, medium, and high.

New Student

• Population: 1, for low (instead of 20%)
• Frequency: 3, for high
• Criticality: 3, for high
• Total Engagement Volume: 9 (1 x 3 x 3 = 9)

Returning Student

• Population: 3, for high (instead of 80%)
• Frequency: 2, for medium
• Criticality: 1, for low
• Total Engagement Volume: 6 (3 x 2 x 1 = 6)

You can show this work in your spreadsheet grid.

You can also work everything into a percentage.

• New Student engagement volume: 9
• Returning Student engagement volume: 6
• Total: 15 (9 + 6)

New Student percentage = 9/15 = 60%

Returning Student percentage = 6/15 = 40%

But this is fake math!

Some of you are thinking about this fake math and screaming, “THIS ISN’T REAL!!!”

You’re right. And that’s ok. It’s ok that it’s not real math because we’re not doing real math. We’re creating one blended metric using several (likely) qualitative metrics.

As long as you clearly explain where the numbers come from, what you’re doing, and why, it’s ok that it’s fake math. The point is to sum up a direction in one number. So you can prioritize. We aren’t plotting navigation and thrust to get a rocket of kittens safely to the moon. This is just directional.

Of course, you should never prioritize alone. This is a group activity. You should be collaborating.

Collaborating on prioritization

I’ve become a huge fan of collaborative workshops and working sessions. The best and worst thing about analyzing something in a room full of stakeholders is getting everyone to agree.

When you work through prioritization exercises together, in a group, then the group as a whole discusses what is important and works through the analysis together. When you walk out, everyone in the room understands where the decision came from.

And, three weeks from now when you’re evaluating call-to-action placement on a wireframe, everyone understands why New Student’s button should be more prominent than Returning Student’s button. More importantly, if you’re looking at the wireframe, and you know that decision is wrong, then you have a clear trigger to spend some conversation re-evaluating your priorities. Is it different just in this screen? Or should it be different in more places?

Visualizing prioritization

When working through prioritization in a group, or even when documenting outcomes, it’s useful to visualize the prioritization, so people can see the values and understand them without needing to think about them.

There are a few ways I’ve done this in the past (and this is by no means comprehensive).

Big numbers

Embiggening numbers makes them easy to see.

This is the easiest way to communicate a blended metric like engagement volume. A big, fat number. If you create a single, blended metric out of multiple criteria, you can arrive at one, big, fat number for each persona. The persona with the biggest number is the most important persona.

Percentage bar

What personas are most important? Ms. Peacock and Mr. Green!

Using a blended engagement volume, score? Add all of the scores up and calculate the percentage of the total each persona uses. Visualize this using a colored bar where each persona’s color represents their percentage. The persona’s with the largest portions of the bar are more important.

I like percentage bars instead of pie charts because they take up less room. Unfortunately, they’re not as easy to understand at first glance. Your audience will need a quick lesson in what’s going on.

DNA Strip

Use DNA strips to compare personas across large numbers of criteria. In this example, Orange Olivia was the persona the client was focused on. Green Gary was the persona that was actually more important according to the criteria the client said was important. When the client could see that their important criteria did not match their perception of the important persona, they were able to change their perception of who the target persona should be.

If you’re not using a blended score, and you still want to compare persona importance across multiple attributes or criteria, you can use a DNA Strip to visualize criteria scores. A DNA strip uses darkness to communicate value instead of raw numbers. Higher numbers have darker colors.

I really dig DNA strips because they just look bad ass.

Like the percentage bar, it takes a teensy lesson on what it means, but it’s pretty intuitive after that. The darker it is, the more important it is, so it’s easy to read at a glance, and easy to visually compare personas side-by-side.

One other advantage of the DNA strip is that you can reveal that some personas may be more important in one area of the system while other personas are more important in other areas.

Spider graphs

With spider graphs, the more color a persona has, the more important they are. Like the DNA strip, you can show how different personas might be more important for different parts of your system.

I pretty much hate spider graphs. They’re like the PowerPoint of graphs. They give you a lot of information in a hard to read format. But, I admit, they do look official and academic, and sometimes that kind of presentation instills additional confidence.

Use the right visualization for your project

From project to project and audience to audience, you will probably need to use different visualization techniques. Remember your goal is to communicate the decision in a useful way, not to look pretty. No matter how you choose to visualize your prioritization, pretty pictures won’t be enough.

Get blue in the face

Like any other guiding principle for your project, you need to remind stakeholders every single time it comes up. The only thing more important than identifying your priorities is remembering your priorities. Repeat them over and over.

Let’s say you’re an interior designer. You would conduct a review something like this:

• We prioritized Dexter the Killer over Nancy the Dealer when we designed this page. (reminder!)
• We placed the butcher knife on the counter, because we want everything to be extra easy for Dexter the Killer to find. Nancy’s baggies are hidden in the drawer because she’s not as important here. (priority: Dexter!)
• We spattered the walls with blood because that’s what Dexter the Killer would like, and we prioritized the design for him. Nancy’s too busy being paranoid to care about blood splatter, anyway. (rationale!)

You’re not repeating the priorities because your stakeholders can’t remember anything. Remember when we talked about headspace 1000 words ago? Your stakeholders are busy. They make lots of decisions. They focus on lots of things and ignore lots of others. The only thing more important than identifying your priorities is remembering your priorities.

The only thing more important than identifying your priorities is remembering your priorities.

The only thing more important than identifying your priorities is remembering your priorities.

Remembering your priorities means you can use less headspace for some decisions, so you can free up headspace for making more difficult decisions.

So, mention your priorities every time. Say them until you’re blue in the face. Help the team remember the priorities that the team identified. You’ll streamline the decision process, so you can focus on the really tough bits.

Once your team has prioritized the personas, it’s much easier to make sure your design is always addressing the right user’s needs in the right way.